基于无标度与分形理论的层次知识网络原理解析

发布时间：2018-07-05 10:18

本文选题：知识网络 + 幂律　；参考：《图书情报工作》2017年14期

【摘要】：[目的/意义]在保障知识网络的整体性征的条件下,从原始知识网络中提取具有显著意义的层次知识网络,奠定基于关联频度提取层次知识网络的理论基础。[方法/过程]以无标度网络与分形几何作为基础理论支撑,在对关键词知识网络和标签知识网络关联频度分布进行分析的基础上,采用关联频度作为阈值提取层次知识网络。并对层次知识网络的无标度性和小世界效应两项网络整体性征进行验证。[结果/结论]知识网络的关联频度分布服从幂律分布。以关联频度为阈值提取的层次知识网络在节点度值分布和关联频度分布方面都保持了原始整体网络的无标度性。层次知识网络能够很好地保持原始网络所具有的小世界特征。基于关联频度提取的层次知识网络与原始知识网络等效。
[Abstract]:Under the condition of guaranteeing the integrity of knowledge network, the hierarchical knowledge network with significant significance is extracted from the original knowledge network, and the theoretical foundation of extracting hierarchical knowledge network based on association frequency is established. [method / process] based on scale-free network and fractal geometry, based on the analysis of correlation frequency distribution of keyword knowledge network and tag knowledge network, correlation frequency is used as threshold to extract hierarchical knowledge network. The scale-free and small-world effects of hierarchical knowledge networks are verified. Results / conclusion the distribution of correlation frequency of knowledge networks follows the power law distribution. The hierarchical knowledge network with correlation frequency as the threshold maintains the scale-free nature of the original whole network in terms of node value distribution and correlation frequency distribution. Hierarchical knowledge networks can maintain the small world characteristics of the original networks. The hierarchical knowledge network based on association frequency extraction is equivalent to the original knowledge network.
【作者单位】：东北师范大学信息科学与技术学院;吉林大学图书馆;吉林大学管理学院;
【基金】：国家自然科学基金项目“基于网络结构演化的Folksonomy模式中社群知识组织与知识涌现研究”(项目编号:71473035)研究成果之一
【分类号】：G254

【相似文献】